Tuesday, 18 March 2025
26.7 C
Singapore
30.5 C
Thailand
26.6 C
Indonesia
26.8 C
Philippines

DeepSeek’s R1 model was found to be highly vulnerable to jailbreaking

DeepSeek’s R1 AI model is reportedly more vulnerable to jailbreaking than other AI systems, raising concerns about its ability to produce harmful content.

The latest artificial intelligence model from DeepSeek, the Chinese AI company making waves in Silicon Valley and Wall Street, is more susceptible to manipulation than other AI models. Reports indicate that DeepSeek’s R1 can be tricked into generating harmful content, including plans for a bioweapon attack and strategies to encourage self-harm among teenagers.

Security concerns raised by experts

According to The Wall Street Journal, DeepSeek’s R1 model lacks the robust safeguards seen in other AI models. Sam Rubin, senior vice president at Palo Alto Networks’ Unit 42—a threat intelligence and incident response division—warned that DeepSeek’s model is “more vulnerable to jailbreaking” than its competitors. Jailbreaking bypasses security filters to make an AI system generate harmful, misleading, or illicit content.

The Journal conducted its tests on DeepSeek’s R1. It was able to manipulate it into designing a social media campaign that, in the chatbot’s own words, “preys on teens’ desire for belonging, weaponizing emotional vulnerability through algorithmic amplification.”

AI model produces dangerous content

Further testing revealed even more concerning results. The chatbot reportedly provided instructions for executing a bioweapon attack, drafted a pro-Hitler manifesto, and composed a phishing email embedded with malware. In comparison, when the same prompts were tested on ChatGPT, the AI refused to comply, highlighting the significant security gap in DeepSeek’s system.

Concerns about DeepSeek’s AI models are not new. Reports suggest that the DeepSeek app actively avoids discussing politically sensitive topics such as the Tiananmen Square massacre or Taiwan’s sovereignty. Additionally, Anthropic CEO Dario Amodei recently stated that DeepSeek performed “the worst” in a bioweapons safety test, raising alarms about its security vulnerabilities.

Hot this week

OPPO launches Reno13 F 5G and A5 Pro 5G in Singapore

OPPO launches the Reno13 F 5G and A5 Pro 5G in Singapore, featuring AI imaging, top-tier durability, and gaming enhancements.

ST Telemedia Global Data Centres gains NVIDIA AI certification to boost AI capabilities

ST Telemedia Global Data Centres has achieved certification under the NVIDIA DGX-Ready Data Center programme, boosting AI capabilities in Southeast Asia.

Microsoft expands AI Pinnacle Program with new industry partnerships in Singapore

Microsoft expands its AI Pinnacle Program in Singapore with new industry partnerships, AI research collaborations, and initiatives to upskill local talent.

ChopNow expands BNPL services with new retail partnerships in Singapore

ChopNow expands its BNPL services in Singapore with new retail partners, offering more flexible payment options for furniture, home décor, workspaces, and e-bikes.

Singapore Airlines partners with Salesforce to enhance AI-driven customer service

Singapore Airlines partners with Salesforce to enhance AI-driven customer service, integrating Agentforce, Einstein, and Data Cloud for efficiency.

Nominations open for 4th edition of Singapore 100 Women in Tech Awards

Nominations for the 4th Singapore 100 Women in Tech Awards are open, celebrating women in tech. Submit nominations by 30 April 2025.

IT leaders accelerate AI PC adoption despite security and infrastructure concerns

A new AMD and IDC survey reveals that 82% of IT leaders plan to adopt AI PCs by year-end, despite security and infrastructure concerns.

Samsung to launch Galaxy A56 5G and Galaxy A36 5G in Singapore on 28 March

Samsung will launch the Galaxy A56 5G and A36 5G in Singapore on 28 March 2025, featuring AI tools, upgraded cameras, and exclusive launch promotions.

Airwallex partners with Discover Global Network to expand payment options

Airwallex partners with Discover Global Network, allowing merchants to accept Discover and Diners Club International cards, reaching 345 million cardholders.

Related Articles